Numerical caustics

A linear dispersive mechanism for error focusing in polychromatic solutions is identified. This local error pileup corresponds to the existence of spurious caustics, which are allowed by the dispersive nature of the numerical error. ¿From the mathematical point of view, spurious caustics are related to extrema of the numerical group velocity. Several popular schemes are analyzed and are shown to admit spurious caustics. It isalso observed that caustic-free schemes can be defined, like the Crank-Nicolson scheme.

We present mathematics for the construction of presentations using Excel for Ibn Sahl's lens and also Huygen's lens. These methods are based onf Fermat's principle. Huygen's lens can have a spherical frontal surface of a Hyperbolic one. Sahl has both surfaces hyperbolic. This is an advantage in the calculation of the Caustics

In this paper, Software is presented for teaching through interactive demonstrations about caustics on rainbows and lenses. At first, we explore the caustics on rainbows since they are amenable to analytic calculations. We find 4 kinds of caustics: 2 internal (of first refraction and reflection) and 2 externals (second refraction and exciting from the back of the drop, and 2nd refraction emerging on the front). We examine caustics of lenses constructed by two spherical surfaces. We explore the ray diagrams and wave fronts and give methods for finding the caustics. We also examine the case of thick lens and the caustics due to reflection

A linear dispersive mechanism leading to a burst in the L∞ norm of the error in numerical simulation of polychromatic solutions is identified. This local error pile-up corresponds to the existence of spurious caustics, which are allowed by the dispersive nature of the numerical error. From the mathematical point of view, spurious caustics are related to extrema of the numerical group velocity and are physically associated to interactions between rays defined by the characteristic lines of the discrete system. This paper extends our previous work about classical schemes to dispersion-relation preserving schemes.

In computer graphics, it is often an advantage to calculate refractions directly, especially when the application is time-critical or when line graphics have to be displayed. We specify efficient formulas and parametric equations for the refraction on straight lines and planes. Furthermore, we develop a general theory of refractions, with reflections as a special case. In the plane case, all refracted rays are normal to a characteristic conic section. We investigate the relation of this conic section and the diacaustic curve. Using this, we can deduce properties of reciprocal refraction and a virtual object transformation that makes it possible to produce 2D-refraction images with additional depth information. In the three-dimensional case, we investigate the counter image of a straight line. It is a very special ruled surface of order four. This yields results on the order of the refrax of algebraic curves and on the shading of refracted polygons. Finally, we provide a formula for the diacaustic of a circle.

In computer graphics, it is often an advantage to calculate refractions directly, especially when the application is time-critical or when line graphics have to be displayed. We specify efficient formulas and parametric equations for the refraction on straight lines and planes. Furthermore, we develop a general theory of refractions, with reflections as a special case. In the plane case, all refracted rays are normal to a characteristic conic section. We investigate the relation of this conic section and the diacaustic curve. Using this, we can deduce properties of reciprocal refraction and a virtual object transformation that makes it possible to produce 2D-refraction images with additional depth information. In the three-dimensional case, we investigate the counter image of a straight line. It is a very special ruled surface of order four. This yields results on the order of the refrax of algebraic curves and on the shading of refracted polygons. Finally, we provide a formula for the diacaustic of a circle.

A linear dispersive mechanism for error focusing in polychromatic solutions is identified. This local error pileup corresponds to the existence of spurious caustics, which are allowed by the dispersive nature of the numerical error. ¿From the mathematical point of view, spurious caustics are related to extrema of the numerical group velocity. Several popular schemes are analyzed and are shown to admit spurious caustics. It isalso observed that caustic-free schemes can be defined, like the Crank-Nicolson scheme.

In this paper, Software is presented for teaching through interactive demonstrations about caustics on rainbows and lenses. At first, we explore the caustics on rainbows since they are amenable to analytic calculations. We find 4 kinds of caustics: 2 internal (of first refraction and reflection) and 2 externals (second refraction and exciting from the back of the drop, and 2nd refraction emerging on the front). We examine caustics of lenses constructed by two spherical surfaces. We explore the ray diagrams and wave fronts and give methods for finding the caustics. We also examine the case of thick lens and the caustics due to reflection.  We also examine the Huygens Lens and the Ibn Sahl lens

Fermat's principle is considered as a unifying concept. It is usually presented erroneously as a 'least time principle'. In this article we present some software that shows cases of maxima and minima and the application of Fermat's principle to the problem of focusing in lenses.

In this paper, Software is presented for teaching through interactive demonstrations about lenses. At first we explore lenses constructed by two spherical surfaces. We explore the ray diagrams and wave fronts. Then there is a page for understanding the thick lens model. We introduce a step by step procedure to find the focal length and find the principal planes and finally the use of the focal length and principal points to construct the image. There is a page for finding the position of the image not by the formula but by the method we use on an actual experiment: We move the screen back and forth until we can get the sharpest possible image. This is done by finding the minimum of a standard deviation of the position of the rays for a given position of the screen. Then there is a simulation of an experiment for finding the focal length. This uses a macro to simulate the finding of several image points b for several object points a. These values are used first in the graphical representation of the image point as a function of b and the image points as a function of a. With suitable least square fits we get two lines with parameters that give values for the focal length and principal plane. Then there is a simulation of two experiments of finding the focal length of a lens. The spreadsheet calculates the distance b vs a, the image y, and there ar graphs of y as a function of a and y as a function of b from which we find 1) a hyperbolic fit for y vs a and a linear fit for y vs b from which we calculate the focal distance, 2) it calculates 1/a and 1/b and then finds a linear fit and a parabolic fit for the data. Also we get the same parameters by finding the cuts of lines uniting the point (a,0) and (0,b).. 3)there is a plot of a+b vs a and then the points are fitted with a hyperbola whose asymptotes give the sum of focal length and principal planes. Then there is a page where we can see two lenses for which the shape can change to have a perfect focusing at a given distance. These two lenses are based on Huygens' ideas (a preprint exists in https://www.researchgate.net/publication/322302527_Spherical_lenses_and_Huygens'_lens?_sg=afv5RV2zzgNXkpkwD-GkA5bFhOCe_gUk-zc1hKeFRQ8GE7IYfzNWVjzoHdibFtsUYG-F0-MzCvhn144VnYj5LyqZV817ao1RdVitW9qB.KOs9BSuYkYtrji9qHmqB8_fOUF4LBrFaixxy-g-zy29AdDNZ3VXnq8Z8Fr-dsjksZBLnZtOIKiv7JkiQSzxTEQ)

A linear dispersive mechanism leading to a burst in the L ∞ norm of the error in numerical simulation of polychromatic solutions is identified. This local error pile-up corresponds to the existence of spurious caustics, which are allowed by the dispersive nature of the numerical error. From ...

A linear dispersive mechanism for error focusing in polychromatic solutions is identified. This local error pile-up corresponds to the existence of spurious caustics, which are allowed by the dispersive nature of the numerical error. ¿From the mathematical point of view, spurious caustics are related to extrema of the numerical group velocity. Several popular schemes are analyzed and are shown to admit spurious caustics. It isalso observed that caustic-free schemes can be defined, like the Crank-Nicolson scheme.